Release of Information and Follow-on Actions after Defects were Found on 26 MRT Trains

Mr Sitoh Yih Pin asked the Minister for Transport whether Kawasaki Heavy Industries and CSR Sifang are liable to compensate LTA and/or SMRT for any loss of use and/or profits (or any other damages suffered) for the downtime of the MRT trains sent back to the manufacturer's factory for rectification in addition to the shipping and defect rectification costs.

Mr Sitoh Yih Pin asked the Minister for Transport (a) whether the hairline cracks found on the car-body bolster of the affected MRT trains were inherent defects from the time LTA and/or SMRT accepted delivery of the trains in 2011 or defects that were manifested through use and subsequently discovered in July 2013; (b) if the hairline cracks were inherent defects, why were the defects not discovered through earlier routine checks; and (c) if these were defects manifested through use, whether LTA engineers expect the hairline cracks to worsen through continued use.

Then, I shall come to it when the turn comes up.

The Land Transport Authority (LTA) did not detect any cracks when it took possession of the trains and when it was testing them in 2011. The hairline cracks appeared only about two years after they were put into operation. It is not uncommon for new trains to develop minor defects after some time, because they are subject to different levels of stress during testing and when they are deployed for actual operations.

LTA did not publicise the hairline cracks for three reasons. First, they did not pose any safety risk to commuters. In fact, I wanted to reply to Question No 12 so that I can elaborate on this point, but we shall come to it later.

Second, the manufacturer took immediate and full responsibility for the defects, committed to fully replace the bolsters with a new set made in Japan, weld them to the new car bodies and re-assemble the trains in China. All these remedial actions would be fully paid for by the manufacturer.

Third, the manufacturer accepted LTA's replacement work schedule which is deliberately formulated to ensure that train services and capacity levels are not affected by this incident.

Had any of these factors not been satisfactorily dealt with, LTA would have publicised the defects. Meanwhile, we continue to monitor the affected trains closely, including those which had been rectified by the manufacturer. Whether there is a case to pursue compensation, I leave it to the LTA Board to assess.

Madam, I thank the Minister for his response. I would like to ask two questions. Firstly, does this series of events, in any way, affect our plans to increase our capacity on the North-South East-West lines? Secondly, it was reported in The Straits Times that there was a freon gas leak yesterday in one of our trains. Can the Minister provide more information to the House, please?

We handled the defective bolsters in such a way that we make sure that the planned capacity expansion of the North-South East-West line will not be affected. As Members know, 26 trains were found to have defective bolsters. Had the manufacturer wanted LTA to return all the 26 trains in one go, then our plan to expand the capacity of the North-South East-West lines would have been affected.

That is why, in 2013, LTA worked out a schedule to which the manufacturer accepted that they will only take one train at a time. In that way, the capacity will not be affected too much because all operators of train service would allow for at least a 10% buffer. For example, for North-South East-West lines, although we have 140-plus trains, only 124 are fully launched at peak hours. We do not need more than 124. Our plan was to expand 124 deployed trains to about more than 150 by the year 2019; which requires the North-South East-West line to expand its fleet from 140- to 170-plus. All the new trains will continue to come. But because we would only release one train at a time for the manufacturer to fix the bolsters, it does not affect our buffer and, therefore, it does not affect our planned capacity expansion of the North-South East-West lines.

To reiterate the point, this is also the reason why the whole process of rectifying, seemingly, takes such a long time. Actually, to fix the bolsters only requires about three to four months. It involves the following activities. We send the trains back. The manufacturer has to dismantle the trains, remove the bolsters. The bolsters are actually welded to the car body. Meanwhile, a new set of bolsters are being manufactured, this time not in China but in Japan. So, they have to be sent over to China and then to be welded to brand new car bodies and the manufacturer has to re-assemble all the engines and so on, to the train. The whole process would take about four months, including testing. Having completed this whole re-assembly, you have to test the trains to make sure they are safe and functioning as required.

But because each time we can only release one train, so we have to queue up the trains and we have a work schedule that stretches all the way to beyond 2020. Recently, we sat down with the manufacturer and asked to speed this up a bit. The possibility is that we can release two trains at a time in the later part of next year, when more new trains come in. We are working on that basis. With that, we could speed up the whole process to complete by 2019.

The query on freon gas. This has nothing to do with the new Kawasaki trains, but yes, it happened yesterday. It is an old train. This is the first-generation Kawasaki train; not this current generation, which is having the bolsters problem. The train and air-con compressors are all 20-odd years old by now. That is the trouble with old parts. Sometimes, little problems happen. Preliminary investigation suggests that there was a leakage of the oil from the compressor which resulted in overheating of the compressor. Subsequently, that caused the evaporation of the freon, which was then released.

We are still checking. This morning, the engineers are going to pull apart the whole compressor to see what caused the leakage from the compressor shaft seal. There is a seal which keeps the oil in place. We need to know why it failed. There is a three-weekly check on the compressor as part and parcel of routine maintenance and the last time we checked on this particular compressor was only about 10 days ago, in the early part of August. So, we need to know what is the reason for this incident.

There is a lesson here, even though it was about train. Old assets are not the same as new assets and, therefore, you have to pay particular attention to maintenance. But the earlier part of the life history of the components is also important. Like a human body. If you look after your body well from Day 1, when you are 40-50 years old, you can probably still maintain pretty good form, but still a 40-year-old is not the same as a 14-year-old, and, likewise, it is the same with trains.

Thank you. Those are good supplementary questions. Let me reply to the Member's questions in the order as I recall.

First, Kawasaki. When they tendered for this, it was Kawasaki with the Chinese company Sifang. I think they are from Qingdao in the Shandong area. They jointly made a bid. But the main contractor, as far as we are concerned, is Kawasaki. In fact, the bulk of the design, how it is laid out, the engines and so on, were all from Kawasaki.

The Chinese partner, essentially, has three roles. First, they produce the bolsters which are made in China; secondly, they produce the car bodies; and thirdly, they assemble. So, all the engines or whatever parts that go into a train – if they are not produced in China, and most of them are not – they will be sent to Qingdao and then Sifang Company will do the assembly, do the testing, and then they will export it to whoever is the customer, in this case, Singapore.

As far as we are concerned, the party that we interact with is Kawasaki. In fact, Kawasaki set up an office here in Singapore for many years now because they were also the manufacturer, designer and producer of our first set of Kawasaki trains; in fact, of our first set of Mass Rapid Transit (MRT) trains. The first lines, North-South East-West lines, were all based on Kawasaki trains. It is a very established, good company that has been around for many, many years.

Their arrangement to do the various assembly work in China is not unique to Kawasaki. In fact, in recent tenders, when we look at the bids, whether it is from Siemens, Bombardier, almost all of them now partner China to do the various things, very similar to what Kawasaki is doing with Sifang, I suppose, partly because of cost competitiveness, in order to bring down the cost of their total bids.

The way Kawasaki dealt with this whole incident – I have not met them but based on reports and discussions I have had with LTA as well as SMRT – I have been very impressed with the way they dealt with this problem. There will always be problems, but at the end of the day, when confronted with a problem, how do you handle it? There are people who will deny, blame others, but in this case, Kawasaki took immediate attention, sent their engineers down, checked, identified the primary cause and, at the same time, worked out how to move from there. It is a very professional response to this problem which I find very impressive.

They took responsibility and fulfilled the warranty period commitments. Although the cracks are not of major safety concern, they decided that a defect is a defect and they will replace it. Although the contract was for the bolster to be made in China, they said, from now on, all the bolsters will be made in Japan. They require the Chinese partner to also produce the new aluminium car bodies, weld it, assemble it and retest it. When they are sent back to Singapore, the defects liability and warranty clock will be restarted. In fact, we have now a very prolonged warranty period for the product. I find that gesture on the part of Kawasaki highly commendable and, if I may say, generous.

The point about other defects of this train, such as battery. This was during the testing period. As we know, when train cars are manufactured or assembled, a lot of testing will be done. During one of those tests, they found that one of the batteries was defective where gases were somehow accumulated in the housing that holds the battery covers. Because of the accumulation of gas, it exploded. Again, Kawasaki took note of that testing result and immediately decided with their Chinese partner to get a new supplier of the batteries. They also redesign the housing, so that this problem will not repeat. Indeed, it was the only incident of this kind which happened.

There were a few cases of glass shattering, not the main glass for the car but on those side windows. Again, they checked. This was also during testing in China. They checked and they found that there was installation error. In other words, the workers could have done it better. Subsequently, when they reminded the workers to do the installation properly, we never heard of such incidents again.

Mdm Speaker, I would like to just clarify on the warranty period. What is the actual period? Also, when we get the replacement car body and bolster, when we reset the clock, is it just resetting the clock on the car body and bolster? The whole machinery and engine are also disassembled and assembled on the new car body. Does the reset of the warranty also apply to the mechanical parts?

The other question I have is: are there any lessons, with the benefit of hindsight, on whether LTA should have given timely information as and when the issue cropped up?

There is a defects liability of one year, then an extended warranty of four years. So, one plus four. In this instance, they decided that they will replace the entire bolsters, including the car body, and then restart the clock. It is fair that the extended warranty only applies to the new bolster and the car bodies. It will be asking too much to require the extended warranty to also cover the engines and other parts, which may not be produced by them. To answer the query, the extended warranty and the defects liability period only apply to the bolsters and the car bodies.

Are there lessons from this incident? Yes, there are. I derived two lessons. First, sometimes, in public communications, providing more information is better than less information. My colleagues who have worked with me now for almost 40 years, know that I always prescribe the principle of over-provide rather than under-provide. This is something that I learnt, especially during the Severe Acute Respiratory Syndrome (SARS) period, in crisis communications. Because at that time, there was a lot of misinformation, a lot of lack of information, a lot of lies going on. And always coming out to share with the people what you know, including what you do not know and admit that "Yes, I do not know at this point in time. I will find out". And when I find out, I will let people know. That is always a better strategy than to try to keep information to your chest.

The second lesson will be, I was struck by how the Kawasaki Manufacturer reacted to this problem: face the problem, get to the bottom of it, own it, not in denial and then go all out to do what is right, even if it means at great cost to the company. That way, you preserve integrity, reputation and also, that is the way one should treat one's customers.

Just one quick question then. The other one I will reserve for the next set of questions. For those trains that are awaiting to be sent back to China and they are in service now, are there any cracks found on the body because the cracks originated from the car body bolster? Are there any cracks found on the body now or developing?

No, the cracks are only on the bolsters. Each car train has two bolsters. This is North-South East-West lines, so each train has six cars. This only happens in some bolsters of some train cars. Were there other cracks elsewhere? No.

I thank the Minister for the answers so far. I just have a quick clarification. Have we ever experienced similar hairline cracks on other trains or models and, if so, what are the models?

The only other times where cracks appeared were in the Bukit Panjang Light Rail Transit (LRT). These are Bombardier LRT trains, different from MRT trains. They found some cracks on the underframe of the cars. This was about last year or the year before. This was before my time. But I recall reading about it. These are old Bombardier trains. They should be coming to 20 years, 18 years or so.

The nature of the cracks is quite different because here, with the Sifang trains, the trains are brand new. I sent to you and then after operation, we found some cracks. Whereas Bombardier trains, they were all right until last year or so, and then there were some cracks on the underframe.

Bombardier also took it quite seriously even though this is now an 18-year-old or 17-year-old product. They sent their chief structural engineer down to take a look, inspect. First, they needed to establish if it was safe to operate, and they confirmed that it was safe to operate. Then, they also decided on what the next step should be. The next step requires regular inspection to see whether it is growing but, more importantly, how to fix the problem.

In that instance, the recommendation of the Bombardier chief structural engineer is to weld it. Therefore, there is no need to send it anywhere. So, they can weld it, in situ, locally in Singapore. They worked out a work schedule on how to do it. I think most of the trains have already been fixed. The other train cars should all be completed by October this year, if I am not mistaken. Meanwhile, the trains are safe to be used.